uranium, nuclear waste, conversion process, dissolution, purification, hydrogen peroxide. * Uranium Chemistry Group, Plasma Technology,. Applied Chemistry ...
http://dx.doi.org/10.17159/2411-9717/2017/v117n8a9
Recovery of uranium from nuclear conversion plant waste by M. Potgieter*†, J.C. Barry*, D.J. van der Westhuizen†, and H.M. Krieg†
The ammonium diuranate (ADU) conversion process that was operated at the Nuclear Energy Corporation of South Africa (Necsa) in the past generated a significant amount of waste containing high concentrations of uranium, which can be re-used if the uranium can be recovered in a useful form. To attain this objective, the composition of the waste material and the amounts of impurities present were determined, followed by an investigation into various methods of uranium dissolution. For dissolution, water as well as different acid types and concentrations were investigated, and the efficiency of each method determined in terms of the uranium recovery as well as the extent of impurities extraction. It was found that the waste material was soluble in HNO3, H2SO4, and HCl as well as water, with a maximum uranium extraction of 98% achieved in 3 M HNO3 in 1 hour at a temperature of 80°C without the addition of an oxidizing agent. The thorium impurity content in relation to uranium was reduced from 8.4% to less than 1% with all acids investigated, as well as water. The most significant reduction in the total impurity content, from 24.3% to 10.8 %, was observed when using water, although this did not result in the highest uranium extraction. 63/-'+ uranium, nuclear waste, conversion process, dissolution, purification, hydrogen peroxide.
21-/',&10/2 The ammonium diuranate (ADU) conversion process operated in South Africa from the 1970s was based on the direct conversion of ADU obtained from different mines (Ponelis, Slabber, and Zimmer, 1986; Ponelis, 1989). This approach differed from those of other countries operating conversion processes at that time in that the feed was not of nuclear grade. Since the ADU was sourced from various South African mines, the composition of the feed to the conversion plant at the Atomic Energy Corporation (AEC) varied. In addition, tail-end distillation of the uranium hexafluoride (UF6) product was used, and as a result of the variability in the ADU a considerable proportion of unwanted elements or impurities, such as sodium (Na), potassium (K), and calcium (Ca), was present during the conversion process (Ponelis, Slabber, and Zimmer, 1986; Ponelis, 1989).
VOLUME 117
* Uranium Chemistry Group, Plasma Technology, Applied Chemistry, Nuclear Energy Corporation of South Africa (Necsa), South Africa. † Membrane Technology Group, Chemical Resource Beneficiation (CRB), North-West University (NWU), South Africa. © The Southern African Institute of Mining and Metallurgy, 2017. ISSN 2225-6253. This paper was first presented at the Hydrometallurgy Conference 2016 ‘Sustainable Hydrometallurgical Extraction of Metals’, 1–3 August 2016, Belmont Mount Nelson Hotel, Cape Town.
785
L
2/ +0+
Basically, the ADU conversion process entailed the conversion of ADU to uranium tetrafluoride (UF4), which was then converted to uranium hexafluoride (UF6) in a fluorine flame reactor (Ponelis, 1989). The final product, thermally stable UF6 gas, was then filtered to remove any foreign particles. Thereafter, the UF6 was frozen out and distilled before being fed to the enrichment plant. Due to the incomplete conversion of UF4 to UF6 in the fluorine flame reactor, a significant amount of solid waste rich in uranium was formed, which accumulated at the bottom of the reactor. The incomplete conversion could have been caused by various factors, including sintering due to the low melting point of UF4, inhomogeneous feed compositions, and the presence of various impurities. A study to determine the effect of alkali metal impurities on the conversion of UF4 to UF6 showed that the presence of these impurities had a significant effect on the sintering of UF4 (Ponelis, 1989). Owing to the high uranium content, the unreacted material is a nuclear liability and needs to be processed to recover a product that can be re-used in possible future conversion activities while reducing the amount of waste that is currently stored in drums. A search of the literature was undertaken to determine whether similar waste exists elsewhere, and how it is handled. Sasahira et al. (2007) mention an ‘ash’ that was formed during fluorination of UO2 to UF6, while Ohashi,
Recovery of uranium from nuclear conversion plant waste Murashita, and Nomura (2014) extracted uranium from UF4 residue and NaF adsorbents originating from conversion activities. It thus became clear that this waste material is unique, in terms of its intrinsic variety in composition as well as the amounts and types of impurities present. In view of the solvent extraction (SX) process that is used to purify uranium (Kumar et al., 2011), the waste material in its current solid form will have to be dissolved to obtain an aqueous solution containing the desired uranium. Furthermore, in order to extract uranium successfully using SX, the fluoride content should be low since fluoride is coextracted by tributyl phosphate (TBP) (Volk, Vakhrushin, and Mamaev, 2000; Coleman, 1966). It can be assumed that a significant amount of the contained uranium is in the UF4 form, since the waste material was generated in the UF4 to UF6 conversion step in the process, and unreacted material would therefore consist of UF4. For the envisioned SX process, provision should be made for the oxidation of insoluble U(IV) to soluble U(VI) during dissolution, possibly by means of the addition of an oxidizing agent. UF4 can be completely solubilized by direct treatment with concentrated nitric acid (HNO3) (Floreancig, 1983, Ohashi, Murashita, and Nomura, 2014). Luk'yanchev and Nikolaev (1963) studied the dissolution of UF4 in sulphuric (H2SO4) and hydrochloric (HCl) acid, and found that the solubility of UF4 increases with increasing HCl concentration, and reaches a maximum in H2SO4 at approximately 3 M. Ohashi, Murashita, and Nomura (2014) used an oxidizing agent (H2O2) during dissolution of UF4 in H2SO4 and HCl as part of a study on the extraction of uranium from fluorine-containing waste. They found that UF4 could be completely dissolved in 1.25 M HCl or 0.3 M H2SO4. The aim of this investigation was, therefore, to dissolve the current waste material and to obtain the highest possible recovery of uranium in solution. Since no reference to similar waste material could be found, and a low solubility was expected due to its sintered nature, the waste material was milled prior to dissolution. The dissolution efficiency was investigated using water as well as different acids (HNO3, H2SO4, and HCl) with and without the addition of H2O2 as an oxidizing agent. Water was included as a washing step prior to dissolution to determine its suitability for removing any water-soluble impurities. Since the waste material is known to contain high levels of fluoride (approx. 20%), different methods will be investigated for removing fluoride subsequent to dissolution but prior to SX, but this falls beyond the scope of the current study. In this paper the focus is mainly on cations in the waste material with concentrations above 0.5 mass%.
3-0(321.*4
The composition of the waste material was determined, focusing on the uranium (U) content and other major constituents that may ultimately influence the SX-based uranium recovery process. A sample of the waste material was dissolved in nitric acid (65 wt% HNO3) in a 1:1 water to
L
786
VOLUME 117
acid ratio. The solution was evaporated to dryness and the residue was re-dissolved in HNO3 at the same concentration to achieve complete dissolution, followed by analysis using inductively coupled plasma-optical emission spectrometry (ICP-OES).
Nitric acid (HNO3, 65 wt%), sulphuric acid (H2SO4, 98 wt%), and hydrochloric acid (HCl, 32 wt%) were obtained from Merck. The dissolution tests were conducted using HNO3 at concentrations of 3 M and 1 M, H2SO4 at 2.5 M and 1.25 M, and HCl at 3 M and 1 M, as well as ultrapure water. Experiments were conducted in the absence and presence of 0.4 mL H2O2 (30 wt%, from Merck) as an oxidizing agent to aid the oxidation of insoluble U(IV) to soluble U(VI). For the solubility measurements, a 1 g sample of the waste material was added to a 100 mL acid solution and heated to 80°C on a hot plate. The temperature was chosen to be high enough to ensure fast reaction kinetics, but low enough to ensure that the solution did not boil, which would cause unwanted loss of solution as well as the evolution of noxious fumes. Each dissolution experiment was continued for 1 hour with continuous stirring using a stirrer bar, to ensure sufficient time for the reaction to proceed, after which the mixture was left to cool to room temperature. The mixture was then centrifuged and decanted to separate the solution from the residue. The residue was washed with distilled water, dried in an oven at 100°C, and weighed. A sample of each solution and residue was analysed using ICP-OES. Using the acid with which the highest uranium recovery was attained, a more comprehensive study was subsequently undertaken in which additional acid concentrations were investigated under identical experimental conditions to the screening investigation described above.
3+,*1+4.2'4'0+&,++0/2
A rough estimate of the composition of the waste material was determined (data not shown) by X-ray fluorescence (XRF). Based on this data, the cations present at concentrations greater than 0.5% by mass were determined using ICP-OES. The results are shown in Table I. Since uranium will be recovered by means of SX, lowering the amount of impurities that may have an influence on the SX process would be an added benefit. From Table I it is clear that Fe (28.0 mg/g), Th (44.5 mg/g), and Ca (50.0 mg/g) are the main impurity elements in the waste material. Currently, there are no clear specifications for the final product material that would result from recovery of uranium. However, it is envisaged that the product would be re-used in a nuclear conversion plant. Therefore, the ASTM standard for uranium ore concentrate (UOC) can be used to specify the maximum allowed levels of impurities of the final product. These are 1% for both calcium and iron and 2.5% for thorium (ASTM, 2013). The focus of the current study was to dissolve the waste material as a preparatory step for future purification by SX, and lowering the levels of these impurities would be a benefit if it can be achieved during this
Recovery of uranium from nuclear conversion plant waste Table I
*3(321+401%4&/2&321-.10/2+44 44(.++40241%34.+134(.13-0.*4 *3(321
/2&321-.10/2 $(""#
1.2'.-' '3 0.10/2
*3(321
/2&321-.10/24 $(""#
1.2'.-' '3 0.10/2
Al
11.0
0.14
Ca
50.0
0.57
Fe
28.0
0.14
K
7.5
0.07
Si
8.0
0.00
Na
8.0
0.00
Th
44.5
0.35
U
487.0
5.80
step. To facilitate the discussion on impurities, the content of each impurity was expressed as a percentage of the uranium according to the relationship:
together with the rest of the waste material not converted to UF6. Since the residue was re-fed to the flame reactor to reduce the amount of waste, thorium as well as other impurities were further concentrated in relation to uranium.
[1]
where i and U refer to the impurity and uranium respectively. The values are presented in Table II. As mentioned previously, it can be assumed that most of the impurities originate from the feed material to the conversion process. Accordingly, the high thorium content resulted from the non-volatility of the formed ThF4, which caused it to settle at to the bottom of the flame reactor
Table II
( ,-014&/21321+4-3*.10 341/4,-.20,( *3(321
To identify the most promising lixiviant for the extraction of uranium from the waste material, a series of screening experiments was conducted using nitric, sulphuric, and hydrochloric acids at two concentrations, with and without addition of H2O2. Another objective was to determine whether the more water-soluble impurities could be removed during this step, even if uranium dissolution was not achieved. The results for the screening experiments in the absence of H2O2 are presented in Figure 1. The percentage dissolution of each element was calculated using the following relationship:
( ,-014
Ca
9.3
Fe
5.4
Th
8.4
Other (Al, K, Na, Si)
1.2
[2] where E represents the specific element investigated. It can be seen that the uranium dissolution was above 78% for all the acids, and 82% for water, with the highest
VOLUME 117
787
L
0",-34))3&14/)4'0))3-3214*00 0.21+4/241%34'0++/*,10/24/)4 .-0/,+43*3(321+40241%34.+32&34/)4H2O2
Recovery of uranium from nuclear conversion plant waste uranium recovery (98%) observed in 3 M HNO3. It is noteworthy that the thorium concentration in solution remained below 5% in all instances. The solubility of calcium, however, was significantly different in in different acids, and very limited in water. Although the calcium content in the waste material could be due to a mixture of various compounds, the behaviour observed could be explained by the presence of calcium sulphate dihydrate (CaSO4.2H2O), the solubility of which is 0.222 g/100 mL in hot water and 0.241 g/ 100 mL in cold water (CRC, 1976). Other calcium compounds show either very high or very low solubilities in water. The solubility of CaSO4.2H2O increases with an increase in acid concentration up to a maximum value (3.5 to 4.5 M for HNO3, 2.5 to 3 M for HCl, and 1.5 M for H2SO4), and then decreases if the acid concentration is increased further (Vershkova et al., 2003; Cameron and Breazeale, 1903; Ling and Demopoulos, 2004; Li and Demopoulos, 2005). For H2SO4, the initial increase in solubility of CaSO4.2H2O has been attributed to the influence of the second dissociation constant of H2SO4, due to the higher solubility in the presence of HSO4– as well as to the increase in ionic strength in the presence of more H2SO4 (Marshall and Jones, 1966). The decrease in solubility at higher concentrations is believed to be due to a combination of effects which include changes in activity coefficient as well as salting-out due to the common-ion effect (Ling and Demopoulos, 2004; Calmanovici et al., 1993). The increasing and decreasing solubility in HNO3 was not clearly explained by Vershkova et al. (2003); however, Zhang et al. (2011) observed a decrease in the solubility of CaSO4.2H2O in the order HNO3 > H3PO4 > H2SO4 > Ca(NO3)2, also without offering an explanation, but which agrees with the results we obtained. A possible explanation for the behaviour in HCl was given by Li and Demopoulos (2005), who ascribed the initial increase in solubility to the presence of HSO4– formed during dissolution of CaSO4.2H2O, and the decrease at higher HCl concentrations to the influence of ion activity coefficients. While the values in Figure 1 cannot be compared directly with those from the literature, due to the complexity of this waste material, the above does explain the higher calcium extractions at higher HNO3 and H2SO4 concentrations. The lower calcium extraction at higher HCl concentration may indicate that the maximum CaSO4.2H2O solubility was at a HCl concentration below 3 M. The purities of the final uranium materials (calculated using Equation [1]) with regard to the three most prevalent impurities (Ca, Fe, and Th) are given in Table III. It is clear that both calcium (in acid) as well as iron (all acids and water) are nearly completely extracted during dissolution (see Table II for original concentrations). Some concentrations are higher than the original concentrations since the change in the amount of uranium extracted (
